“Mengyuan Zhao 2016-01-25”版本间的差异
来自cslt Wiki
（以“* Do some experiment on "chain" model, and got some conclusion: # "chain" model is not sensitive with the number of output layer (pdf_num); # post layer need more co...”为内容创建页面） |
（没有差异）
|
2016年1月24日 (日) 09:57的最后版本
- Do some experiment on "chain" model, and got some conclusion:
- "chain" model is not sensitive with the number of output layer (pdf_num);
- post layer need more context than prior layer
- Start 11000h "chain" model training for sinovoice.
